How to filter pandas dataframe using lambda function with regular expression to extract date

Say I have a pandas dataframe like below (with millions of rows) –

data = {'s3_path': ['s3://mybucket/date=2023-10-26/f1.txt', 's3://mybucket/date=2023-10-25/f2.txt', 's3://mybucket/date=2023-10-24/f3.txt', 's3://mybucket/date=2023-10-23/f4.txt']}
df = pd.DataFrame(data)

I want to filter S3 paths that are before 2023-10-24. What would be an efficient way to do that in pandas? Not knowing a lot about pandas, what I can think of is below, but it is not still complete:

date_cutoff_str = '2023-10-24'
date_cutoff_obj = datetime.strptime(date_cutoff_str, '%Y-%m-%d')

def is_before(cur_date, cutoff_date):
    if cur_date < cutoff_date:
        True
    return False

date_regex_pattern = r'\d{4}-\d{2}-\d{2}'
filtered_df = df.apply(is_before, cur_date=how_do_i_get_regex_value_here, cutoff_date=date_cutoff_obj)

Any suggestion/answer would be greatly appreciated. Thank you.

>Solution :

A possible solution, which uses regex to extract the dates as strings, and then converts them to datetime to compare with 2023-10-24:

df.loc[
    pd.to_datetime(
        df['s3_path'].str.extract(r'date=(\d{4}-\d{2}-\d{2})')[0])
    .lt('2023-10-24')]

Output:

                                s3_path
3  s3://mybucket/date=2023-10-23/f4.txt